Investigating the Effects of Prescribing Different Sea Surface Temperature Reconstructions on the Climatic Conditions During the Last Glacial Maximum by Simulation

In this study, we perform simulations with the ECHAM6.3-HAM2.3 climate model with two prescribed different reconstructions of sea surface temperatures (SST) for the Last Glacial Maximum (LGM) as boundary conditions. While one of the datasets suggests a global cooling of 4.1°C (GLOMAP; Paul et al., 2021), the other suggests a much stronger cooling of 6.1°C (Tierney et al., 2020) during the LGM compared to pre-industrial climate conditions. The comparison of our simulation results to LGM land surface temperatures reconstructed based on noble gas concentrations in groundwater (Seltzer et al., 2021) does not indicate clearly which SST dataset results in a better agreement between our simulation results and observational data. The colder SST dataset provided by Tierney et al. (2020) leads to a good agreement in low latitudes and systematically too cold temperatures in mid-latitudes along the glacier margins. Prescribing the GLOMAP SST for our simulations results in too warm temperatures in low latitudes, but a good agreement in the mid-latitudes. For further assessment, we also compare for both SST datasets the simulated mineral dust deposition in the Southern Hemisphere to observational data (Kohfeld et al., 2013). While GLOMAP SST result in a strong overrepresentation of Australian mineral dust deposited over Antarctica, the SST provided by Tierney et al. (2020) indicate Patagonia to be the dominant dust source during the LGM in terms of deposition over Antarctica with minor contributions from Australia and South Africa. Such dominant Patagonian dust source is in agreement with geochemical data from East Antarctic ice cores (Basile et al., 1997; Delmonte et al., 2008). The differences in individual source contributions can be traced back to changes in the meteorological conditions in the source regions, including vegetation, wind speed and precipitation. Both SST datasets result in a different zonal atmospheric circulation in terms of shape, waviness and velocity, which is caused by the different temperature gradients between the pole and the low latitudes.